High-variability exposure

Study many different instances of each category to build a discrimination that generalizes.

Why it works

Perceptual learning is not memorizing specific examples — it is extracting the invariant structure that defines a category across its varied surface forms. Exposure to high-variability instances forces the perceptual system to ignore irrelevant variation and attend to the features that actually signal category membership. Low-variability training tends to produce recognition that works only for the specific instances encountered.

How to do it

1. For each category you are learning, collect examples that vary in surface features but share the defining structure.
2. Study them in random order rather than grouped by similarity.
3. When you encounter an unfamiliar instance, try to name the category before checking — the mismatch is the learning signal.
4. Actively seek edge cases and atypical exemplars, not just canonical ones.

Evidence

Kellman and colleagues found that perceptual learning modules using varied examples produced reliable improvements in categorization speed and accuracy, with transfer to novel instances not seen during training. This extends a classic finding — Posner and Keele (1968) showed that studying varied distortions lets learners abstract a category prototype that transfers to unseen instances, and Kornell and Bjork (2008) found that interleaving varied exemplars improves inductive category learning over blocked study. (rct)

Most controlled studies are lab-based category learning; transfer to complex real-world expert domains (medicine, law) is promising but less systematically established.

Sources

• Kellman & Garrigan (2009), "Perceptual learning and human expertise," Physics of Life Reviews
• Kellman, P. J., & Garrigan, P. (2009). Perceptual learning and human expertise. Physics of Life Reviews, 6(2), 53-84.
• Posner, M. I., & Keele, S. W. (1968). On the genesis of abstract ideas. Journal of Experimental Psychology, 77(3, Pt.1), 353-363.
• Kornell, N., & Bjork, R. A. (2008). Learning concepts and categories: Is spacing the "enemy of induction"? Psychological Science, 19(6), 585-592.

Common mistake

Training exclusively on textbook exemplars, which teaches recognition of idealized cases but leaves the learner blind to the messy, atypical instances that dominate real-world encounters.